N°35 NOVEMBER 2016
ECONOTE
Societe Generale
Economic and sectoral studies department
POPULATION AGING: RISK OF DEFLATION OR INFLATION?

The world is going through an unprecedented phenomenon in terms
of aging as a result of two forces: falling fertility and rising longevity. All
countries are experiencing this shift, albeit to different extents and with
different timings. Japan is at the forefront of the demographic changes, but
the rest of the advanced world is now following in its footsteps as a lower
birthrate in the past and continuing reductions in mortality begin to bite.

Demographics in the advanced world are now at a turning point.
According to the UN, the combined working-age population of the world’s
advanced economies has probably peaked and is expected to fall by more
than 5% in the next five decades. In China, the working-age population is
expected to collapse by a third by 2065. At the same time, the share of these
countries’ population aged 65 and over will skyrocket.

This paper aims at reviewing the consequences of aging for price
dynamics. There has been scant theoretical and empirical work to date on
the potential relationship between aging and inflation, and the limited
research on this topic is not conclusive. Recent work suggests that fast-aging
countries could confront rising inflation pressures, which challenges the more
common view that an aging population is inherently deflationary.

Over the past twenty years, Japan has endured essentially the same
demographic experience most developed countries are set to face over the
next fifty years, and the country has long been mired in deflation. Currently,
several aging countries are also experiencing historically low inflation or even
deflation. While it is possible that population aging could lead to sustained
inflationary pressures in the future, Japan’s experience suggests, though,
that the opposite could be more likely. Assessing whether Japan’s
experience could provide a predictive model for other economies is clearly an
issue worthy of further research.
OLD AGE DEPENDENCY RATIO
(Ratio of population aged 65+ per 100 population aged 20-64)
90
80
China
70
60
Japan
50
UK
40
France
30
Germany
20
USA
10
Italy
2100
2090
2080
2070
2060
2050
2040
2030
2020
2010
2000
1990
1980
1970
1960
1950
0
Source: UNDP.
Marie-Hélène DUPRAT
+33 1 42 14 16 04
[email protected]
ECONOTE | N°35 – NOVEMBER 2016
Seven years after the global financial crisis, growth in
the advanced world has repeatedly disappointed while
inflation has remained stubbornly below main central
banks’ target. Economists have come up with a variety
of
explanations
to
account for
this
weak
macroeconomic performance, such as excessively high
indebtedness, a gradual drop in potential growth and
the progressive loss of policy room for maneuver, both
fiscal and monetary. But commentators have also
increasingly pinpointed the potential role of
demographic trends in recent macroeconomic
underperformance. Recently, a number of central
bankers and researchers have suggested that there
could be a link between low inflation and population
aging.
growth1, there has been scant theoretical and empirical
work on the potential relationship between aging and
inflation. The impact of aging on inflation dynamics is
ambiguous in theory owing to various conflicting
channels. In this paper, we start by briefly describing
the main elements of the demographic transition the
world is currently experiencing. Next we review the
main theoretical channels through which changes in
demographics can possibly affect inflation dynamics.
We emphasize that the graying of the population could
potentially influence price dynamics through a wideranging set of key macroeconomic variables. We
conclude that it is hard to tell from the theoretical
perspective how changes in demographics can shape
inflation rate behavior.
With a declining population that is aging faster than any
other country on the planet, Japan is at the forefront of
the demographic shift (see Box 1), and it has long been
mired in deflation. There are good reasons to believe
that a rapidly declining and aging population has played
a substantial role in Japan’s decades-long growth
slowdown and deflation trap. Several European nations
are now following in Japan’s demographic footsteps,
having experienced earlier and relatively larger declines
in fertility rates than developed countries such as the
United States. This has led some commentators to
believe that, barring a big rise in immigration, these
countries could slide into a Japanese-like regime.
The unprecedented aging phenomenon2 the world is
now experiencing has emerged as a key medium- to
long-term policy concern given the potentially important
implications for central banks, governments, and the
investment banking world at large. Against this
backdrop, the question of the impact of aging on
inflation is bound to become a high-profile topic for
debate on academic and policy research agendas in
years ahead.
RAPIDLY AGING POPULATIONS…
THE DEMOGRAPHIC TRANSITION
The demographic transition refers to the modernization
process of societies from a pre-modern regime of high
birth and death rates to a post-modern one in which
both birth and death rates are low3. Before the
1
See, for example, International Monetary Fund (2004), The
Global Demographic Transition, World Economic Outlook,
Washington D.C., September. See Chap. III, “How will
Demographic Change Affect the Global Economy”, pp. 137-180.
2
The United Nations (UN) definition for “aging” is populations with
more than 7% elderly at age 65 and above. When populations
have more than 14% and 20% elderly, the UN defines them as
respectively “aged” and “super-aged”.
While economists generally agree that the aging of
populations creates fiscal strains and reduces output
3
See notably Frank Notestein (1945), “Population: The long view”,
in T. Schultz (ed.), Food for the World (Chicago) pp. 36-57; W.S.
Thompson (1929), “Population”, American Journal of Sociology,
34, pp. 959-975.
2
ECONOTE | N°35 – NOVEMBER 2016
transition’s onset, high birth and mortality rates cause
the population to grow only slowly. The demographic
transition begins when the mortality rate declines
(mainly because of increased infant and child survival),
so that the country’s population growth accelerates and
the share of youth (children and young adults) in the
population increases (the “boom” generation). But of
course, the youth “bulge” associated with declining
mortality during the early phase of the demographic
transition is temporary as the “boom” generation
gradually works its way through the age structure.
Later, at an intermediate stage, fertility begins to decline
(partly as a result of educational progress and the
increased use of contraception), reducing the number of
children, and the share of the working-age population
(generally taken to be 15-64) increases. As the workingage cohort produces the lion’s share of a country’s
economic output, per capita income then grows more
rapidly, which is often referred as the “sweet spot” or
the “demographic dividend”. Eventually, though, the
“boom” generation reaches the older ages. This secular
“age wave”, combined with lower fertility and continuing
improvements in old-age mortality, then reduces the
growth rate of the working-age share and speeds
growth of the elderly population, causing the sweet spot
to disappear and ultimately to reverse. During the late
stage of the demographic transition, low mortality and
fertility increase the share of the older population, a
process known as population aging, which raises
dependency and puts pressure on family support
systems and public transfer programs (pensions and
health care).
An unprecedented aging phenomenon is what is
happening now in most developed countries. The
demographic sweet spot the advanced world has been
living through since around 1970, when those born in
the post-war baby boom entered the workforce, is now
coming to an end, as the retirement of the huge baby
boom generation (born from 1946 through 1964),
combined with the effect of a lower birth rate in the past
causes the share of the population of working age to go
into reverse. Simultaneously, the older populations in
the advanced world continue to grow at an
unprecedented rate.
BOX 1. JAPAN'S DEMOGRAPHIC UPHEAVAL
With a population that is declining and aging faster than any country on the planet, Japan is at the forefront of
the demographic transition, and so may provide a laboratory for studying the demographic challenges
awaiting the world.
Japan’s fertility has been below replacement level (that is, 2.1 births per woman) for the past 40 years. The
country’s fertility rate started to decline as early as the 1920s, and then rebounded in the 1940s, before
showing a sharp drop from 1949 onwards, corresponding to the end of the (first) baby boom in Japan. The
birth rate was then roughly flat at around 2 until the mid-1970s, when it began to fall dramatically.
By 2005, Japan’s total fertility rate (TFR) – the average number of children born to each woman in her
childbearing lifespan (15–49 years old) – had dropped to 1.26 births per woman. It then rose slowly, but
steadily, to reach 1.43 in 2013. The fall in the Japanese birth rate can be traced back to many factors
including the spread of contraception, the higher educational level of women, a declining rate of marriage, a
rising age at marriage, the expanding participation of women in the workforce, a lack of childcare facilities, a
non family-friendly work environment, economic stagnation, and the rising cost of child-raising.
Japan’s fertility rate is not, though, the lowest in the OECD. In 2013, with its 1.43 TFR, Japan was doing
4
better than South Korea and Singapore (both 1.19), Hong Kong (1.12) and Germany (1.38) . But the fact that
Japan has had a low fertility rate for a long time has meant that the pace of decrease in the growth rate of its
population has virtually been globally unprecedented.
4
Source: the World DataBank.
3
ECONOTE | N°35 – NOVEMBER 2016
Sustained immigration could have possibly delayed the population decline, but Japan has long been
characterized by a very low level of immigration, mainly owing to very restrictive policies for unskilled
migration. Immigrants accounted for only 1.61% of Japan’s total population in 2015, twice as much as that of
a decade ago, but still very low when compared with many other countries, especially Canada (22%), North
5
America (14% ), and even the UK (13%), Germany (15%), France (12%) or Italy (10%).
Falling fertility, combined with very low levels of immigration has been reflected in a continuous decline of
Japan’s population growth since 1975. Japan’s population growth hit zero in 2011 and then moved into
negative territory.
JAPANESE POPULATION GROWTH
(Annual percent change)
1,5
1,0
0,5
0,0
-0,5
Source: UNDP.
2100
2090
2080
2070
2060
2050
2040
2030
2020
2010
2000
1990
1980
1970
1960
-1,0
Meanwhile, starting from the early 1950s, Japan has registered large gains in longevity, which have produced
not only an older population but also an older workforce, relative to other advanced economies. With
increasing longevity and falling fertility, Japan has experienced both a growing population of older workers
and a shrinking population of younger ones.
At 85, the Japanese life expectancy is now the highest in the world, and the UN reckons that the country's life
expectancy could be 94 by the year 2095. Since the early 1990s, the share of the population aged 65 years
or older has risen sharply, from 12% in 1990 to 25% in 2013; it is set to keep increasing to a projected 40% in
2060. Meanwhile, the share of youth, below the age of 15, fell from 16% in 1995 to 13% in 2013. It is slated to
fall further to a projected 10% in 2060, due to the drop in fertility.
In parallel, the drop in fertility has led, with roughly a 15- to 20-year lag, to a sharp fall in the share of the
population in the working-age group (from 15 to 64 years of age), from a peak of 69.5% in 1995 to 60.8% in
2015.
WORKING-AGE POPULATION (15-64) IN JAPAN
(In percent of total population)
75
70
65
60
55
50
45
40
The share of the working-age population will continue to fall in coming decades to a projected 51% by 2060,
which means that, by the year 2060, each Japanese person of working age will be supporting one person of
6
non-working age, which is a huge "dependency ratio" .
5
But 26% if one adds to the immigrants their US-born children.
6
The dependency ratio is a measure showing the number of dependents aged zero to 14 and over the age of 65, to the working-age
population, aged 15 to 64.
4
ECONOTE | N°35 – NOVEMBER 2016
SLOWING POPULATION GROWTH
Gr7. TOTAL FERTILITY RATE
Gr6. POPULATION GROWTH
(Annual percent change)
3,0
2,5
2,0
World
1,5
More developed
regions
1,0
0,5
Less developed
regions
0,0
1960
1970
1980
1990
2000
2010
2020
2030
2040
2050
2060
2070
2080
2090
2100
-0,5
Source: UNDP.
From the 1960s onwards, the world’s population growth
rate started to decline (from a high of 2.1% a year
between 1965 and 1970 to 1.1% currently) as a result
of a historically unprecedented fall in fertility in virtually
all major areas of the world9. Throughout the late 1960s
and early 1970s, the majority of industrialized countries
went through a rapid fall in fertility that reflected such
factors as contraceptive availability, changing attitudes
about the status of women and the expansion of
women’s opportunities to work outside the home. In
Japan, the UK, the US and all European countries,
fertility is now below the level required for full
replacement of the population in the long run (around
2.1 children per woman, on average)10.
7
World population grew at about 0.05% annually for 10,000 years.
In the 17th and 18th centuries, the population growth rate
increased to about 0.5% annually and reached 0.7% after 1900. In
1900, 1.6 billion people inhabited our planet; by 2000, this figure
had risen to 6.1 billion. The United Nations projects that, by 2100,
the world population will reach 11.5 billion people.
8
For the world as a whole, life expectancy more than doubled
from around 30 years in 1900 to 71 years now, and is projected to
rise to 78 years in 2050-55 and to around 83 years in 2095-2100.
Life expectancy varies considerably across regions, from a low of
61 in Africa to a high of 80 in northern America.
4,0
3,5
Japan
3,0
UK
2,5
Italy
2,0
Spain
1,5
France
1,0
Germany
0,5
USA
2080
2090
2060
2070
2040
2050
2020
2030
2000
2010
1980
1990
1960
1970
1950
0,0
Source: UNDP.
According to the UN’s population projections, the
standard source for demographic estimates, world
population growth is set to continue to fall to a projected
1% in 2020 (the lowest rate since the 1950s) and to
only ¼% by 2080 (which will bring the world back to
17th and 18th century figures). Of course, these global
developments mask significant differences between
countries and regions, especially between developed
areas and less developed regions. For example, Africa
(the fastest-growing major area), the Middle East and
some parts of Asia are still growing at a robust rate,
reflecting their higher fertility rates, while the population
in Europe is hardly growing at all, and population
growth in Japan has dipped into negative territory.
Gr8. POPULATION GROWTH
(Annual percent change)
3,0
2,5
Africa
2,0
China
1,5
Japan
1,0
Southern Asia
0,5
Europe
0,0
-0,5
Germany
-1,0
Northern America
1955
1965
1975
1985
1995
2005
2015
2025
2035
2045
2055
2065
2075
2085
2095
Prior to the twentieth century, world population growth
was slow7. It then rose to a peak of over 2% per annum
in the mid-1960s as a result of unprecedented declines
in mortality rates8. The longevity transition first started in
more advanced countries and then spread to poorer
areas over time. Increased life expectancy reflected
many factors, such as improved life standards, better
nutrition, advances in medical technology, improved
education and, especially in low-income countries,
better sanitation and water systems, along with
important strides in the prevention and treatment of
infectious diseases like HIV, malaria, and polio, among
others.
Source: UNDP.
In the US, population growth is robust in comparison
with that of most other advanced nations, largely thanks
to recent immigrants, who tend to have more children
than residents whose families have been in the country
for several generations. However, in coming decades,
population in America will grow at barely a third of the
annual 1% rate that prevailed from 2000 to 2013.
Between 2020 and 2045, the EU-27’s population is
expected to shrink by around 0.2% per year, with the
decline being particularly pronounced in Italy and
Germany. By 2050, Japan’s population should have
fallen by more than 15%. For the OECD area as a
whole, total population growth is projected to enter
negative territory around 2050.
9
The world fertility rate has dropped to 2.43 births per woman,
down from 4.85 in 1970, with large differences across regions (1.6
in Europe, for example, vs. 4.6 in Africa) and countries (1.2 in
Singapore, for example, vs. 7.6 in Niger).
10
In 2010-2015, fertility was 1.4 children per woman for Japan,
1.55 for China, 1.6 for Europe as a whole (1.28 for Portugal, 1.32
for Spain, 1.39 for Germany, 2 for France), 1.89 for the USA and
1.92 for the UK.
5
ECONOTE | N°35 – NOVEMBER 2016
BOX 2. CAN IMMIGRATION CHANGE GERMANY’S DESTINY?
Germany has succeeded in reversing a decade-long decline in population thanks to four consecutive years of
increasingly high net migration. Net migration to Germany has, in every year since 2011, exceeded 300,000,
with a record high figure of 1,139,000 reached in 2015. Offsetting the decline in Germany’s birth rate, rising
immigration has taken the German population to nearly 82 million people – a level last seen in 2009. In 2015,
Germany registered its biggest annual increase in population (+1.2%) in more than 20 years. The population
growth has been particularly concentrated among those of working age.
Long-term trends, however, continue to point to a demographic decline. The German Federal Statistics Office
warns that not even a million migrants can reverse Germany’s long-term population decline, owing to a
growing birth deficit after 2020 at a time when the large Baby-Boom generation will start to die off. Even if the
birth rate were to rise from 1.4 to 1.6 children per woman, the Statistics Office calculates, overall births would
still decline in the long term. To offset the demographic decline, Germany would need 470,000 immigrants
each year between now and 2040, a flow of migrants which, according to the Statistics Office, is unlikely to be
sustained year after year. The German population is expected to begin to shrink from around the year 2020,
th
as described by the variant “continuity amid weaker immigration” of the 13 coordinated population projection
of the Federal Statistics Office (see the graph below).
Immigration could theoretically slow the developed
world’s decline in population growth. Given migrants’
relatively younger age and higher fertility, immigration
could have a potentially strong and long-lasting effect
on population growth and composition. However,
immigration policy tends to be restrictive in most
advanced countries, partly because of fears that
migration may bring social pressures with it. At present,
only 3.3% of the world’s population lives in countries
other than the one in which people were born11. To
maintain a constant total population, Japan and
European countries would need to raise their
immigration levels multiple times beyond current levels,
which hardly looks feasible. As net immigration into
developed countries will likely remain small overall
11
Nearly 20% of the world’s international migrants are in the
United States, followed by Germany and Russia.
6
ECONOTE | N°35 – NOVEMBER 2016
compared to the size of populations12 and fertility in the
advanced world is unlikely to rise for the foreseeable
future, the current trend of declining populations in the
advanced world appears irreversible.
projected to rise to 4.5% by 2050 and to 8.4% by 2100.
Not only is the world’s population becoming older, the
older are living longer. However, here again, there are
wide differences between countries and regions.
Gr9. NET MIGRATION RATE
G11. POPULATION AGED 65 & ABOVE
(Per 1,000 population)
(Percent of total population)
6
35
30
5
Japan
4
25
UK
3
France
2
Germany
World
20
More developed
regions
15
10
Less developed
regions
USA
1
5
Italy
1950
1960
1970
1980
1990
2000
2010
2020
2030
2040
2050
2060
2070
2080
2090
2100
0
2010
2015
2020
2025
2030
2035
2040
2045
2050
2055
2060
2065
2070
2075
2080
2085
2090
2095
2100
0
Source: UNDP.
RAPIDLY AGING POPULATION
But slowing or declining population growth is not the
only result of the transition from high to low rates of
mortality and fertility. Changes also occur in the age
structure of the population: the world is now
experiencing an unprecedented phenomenon in terms
of aging. The decline in fertility combined with the rise in
life expectancy is causing a dramatic change in the
population age distribution with an increasing share of
old people and a decreasing share of children. The
share of adolescents and young adults is falling in every
region of the world, and in some countries, even the
absolute number of 15- to 24-year-olds is shrinking.
Gr10. POPULATION AGED 15-24
(Percent of total population)
25
20
World
15
10
More developed
regions
5
Less developed
regions
In the advanced world, the cohort of old (65 and over)
and very old (80 and over)13 is growing much stronger
than the cohort of young, with the population of old
people also aging. Population aging is most rapid in
Japan, Western Europe and the United States.
Currently, Japan and Europe have the greatest
proportion of their population aged 60 or over
(respectively 33% and 24%). Between now and 2030,
Japan will age by five years, with the median age
projected to increase to 51.5 years from the current
level of 46.5. Germany, now the second-oldest
population in the world after Japan, will see its median
age increase from 46.2 years today to 48.6 years by
2030. Several other European countries, as well as key
emerging countries such as China, will experience
similar trends14. China is set to face dramatic changes
in the next half-century—a legacy of the one-child
policy—, with a projected tripling of the number of
Chinese people aged 65 and over (from 131 million
currently to about 406 million).
GR12. MEDIAN AGE
60
55
0
1950
1960
1970
1980
1990
2000
2010
2020
2030
2040
2050
2060
2070
2080
2090
2100
Source: UNDP.
Source: UNDP.
50
China
45
Japan
40
12
Net migration should, however, contribute to the expansion or
mitigate the decline of the populations of such countries as the
United States, Germany, Canada, the UK, Australia, Russia and
Italy, which are expected to be the top net receivers of migrants
(with more than 100,000 annually) in the coming decades.
UK
35
Meanwhile, the world’s population aged 60 and over is
rising at an annual rate of 3.26% per year. The share of
people aged 60 and over has risen from 8% of the
global population in 1950 to 12% currently, and is
projected to increase to about 22% by 2050 and to 28%
by 2100. By 2050, all major regions of the world except
Africa will have nearly 25% or more of their populations
aged 60 or over. And the population aged 80 and over,
now accounting for 1.7% of the world’s population, is
30
France
25
Germany
20
USA
15
Italy
2100
2090
2080
2070
2060
2050
2040
2030
2020
2010
2000
1990
1980
1970
1960
1950
10
Source: UNDP.
13
Those aged 80 and older are in the most rapidly growing age
group worldwide. By 2050, the population of centenarians will be
16 times larger than in 2000 (2.2 million compared with 135,000).
14
In Spain, for example, the median age will rise from 43.2 years
currently to 50.1 by 2030; in Italy, from 45.9 years to 50.8; in
Portugal, from 44 years to 50.2; in China, from 43.2 to 48.6 during
the same time period.
7
ECONOTE | N°35 – NOVEMBER 2016
DECLINING WORKING-AGE POPULATION
Population aging combined with lower fertility is having
a profound effect on the growth of the working-age
population. Between now and 2030, all countries,
except a few in the Sub-Saharan African region, will
face either a slower-growing or declining working-age
population15. Japan and Germany, whose working-age
population has been shrinking over the past 15 years,
will continue to exhibit the same trend in the next 15
years. In France, the number of people in the workingage category has grown over the last 15 years, but will
be more or less flat in the next 15 years. Among rich
countries, the UK and the US remain demographically
more fortunate, as the number of working-age people
will continue to grow in 2015-30, but at much slower
rates. In China, the growth rate of the working-age
population has slowed since the late 1980s and it
turned negative in 2015. By 2065, the Chinese workingage population is expected to collapse by a third.
dependency ratio” (the ratio of old people to those of
working age)16. In 2014, Japan had 42 people aged 65
and over for every 100 adults between the ages of 25
and 64. The UN expects that by 2035 this number will
have risen to 69 in Japan. Germany will have 66 old
people for every 100 of working age by 2035 (up from
32 in 2014) and America 44 (up from 22). In China, the
old-age dependency ratio will triple from 12 in 2014 to
36 by 2035.
Gr14. OLD-AGE DEPENDENCY
(Population aged 65 and over per 100 persons aged 15-64)
45
40
35
30
25
20
15
10
5
0
2000
2005
2010
2011
2012
2013
2014
Source: OECD.
Gr13. WORKING AGE POPULATION (15-64 YEARS)
(Percent of total population)
80
75
Japan
70
UK
France
65
Germany
60
USA
2100
2090
2080
2070
2060
2050
2040
2030
2020
2010
2000
1990
1980
1970
50
1960
Italy
China
1950
55
Source: UNDP.
The demographics of the world’s advanced countries
are now at a turning point. According to the UN, the
combined working-age population of the world’s
advanced economies has probably peaked and should
shrink by more than 5% in the next 50 years while the
share of their population over 65 is expected to grow by
80%. By 2050, the working-age population will have
contracted by 28% in Japan, 23% in both Germany and
Italy and 21% in China; in the US, it will have grown by
10% (yet it will be down as a share of total population
from 66% to 60%).
By 2050, it is expected that in many advanced countries
there will be fewer than 2 working-age adults supporting
the burden of one 65+ adult. This will impose huge
demands on the working-age population, if a stable
intergenerational flow of benefits to the elderly
population is to be maintained. It is worth noting,
though, that current measures of dependency which
assume that no one will work after the age of 65 in
coming decades likely overestimate the future financial
burden given the strong assumption that individuals’
behavior and the policy framework will adapt to the
demographic shift. As older workers remain healthy
they will likely want to delay retirement in order to
provide for an adequate standard of living in retirement,
while policymakers will likely extend working lives given
the extra fiscal burden associated with population
aging.
Gr15. OLD AGE DEPENDENCY RATIO
(Ratio of population aged 65+ per 100 population aged 20-64)
90
80
China
70
RISING OLD-AGE DEPENDENCY RATIOS
The falling working-age cohort, combined with the surge
of older persons as a share of total populations, is
reflected in a sharp deterioration in the “old-age
60
Japan
50
UK
40
France
30
Germany
20
USA
10
Italy
2100
2090
2080
2070
2060
2050
2040
2030
2020
2010
2000
1990
1980
See Moody’s (2014), “Population Aging will Dampen Economic
Growth over the Next Two Decades”, Special Comment, August 6.
UN projections used for Moody’s analysis are based on the
assumption of ‘medium fertility’, ‘normal mortality’ and ‘normal
immigration’ (i.e. the future path of international migration is set on
the basis of past international migration estimates and
consideration of the policy stance of each country with regard to
future
international
migration
flows).
See
http://data.un.org/Resources/Methodology/PopDiv.htm#_A._Fertilit
y_assumptions.
1970
15
1960
1950
0
Source: UNDP.
16
The total age dependency ratio is an index that gives a rough
indication of the level of support burden placed on the working-age
group (15-64 years), through comparison of the relative size of the
young-age population and old-age population groups with that of
the working-age group. It is presented as number of dependents
per 100 persons of working age.
8
ECONOTE | N°35 – NOVEMBER 2016
…SHOULD HAVE AN IMPACT ON PRICE
DYNAMICS
According
to conventional
economic
wisdom,
demography does not affect inflation because inflation
is always a monetary phenomenon that can be
controlled by monetary policy. However, controlling
inflation has often proven to be challenging, especially
since the financial crisis of 2008, when most advanced
economies have faced less-than-desired inflation rates
despite ultra-easy monetary policies. This may suggest
that something other than monetary policy is influencing
inflation. Until recently, there had been little research
into how changes in demographics affect inflation – with
most work on aging focusing on the effects on growth
and fiscal sustainability17. But a growing body of
research is now investigating the link between aging
and prices. However, there is no agreement among
economists on whether population aging is inherently
inflationary or deflationary.
This is because declining and aging populations can
potentially influence price dynamics through a
multiplicity of channels, both on the demand side and
on the supply side of the economy, with some
transmission mechanisms being inflationary, others
deflationary, and some having an ambiguous effect, so
that the net effect on prices is by no means
straightforward. A secular change in demographics
leads to an entire transition path that alters saving and
investment behaviors, consumption patterns, the size of
the workforce and, consequently, potential employment
and output growth, the rate of productivity growth, the
returns to factors of production (such as capital and
land) and other key macroeconomic variables in a way
which is not clear a priori. Moreover, the magnitudes of
these various effects depend on many factors, such as
the relative speeds of adjustment of aggregate supply
and aggregate demand, the characteristics of the labor
market, the extent of behavioural responses and the
policy response of governments. At this stage in the
research, the question as to whether population aging
has a deflationary or an inflationary impact does not
have a definitive answer. This is an issue which is
clearly worthy of further academic research.
inflation in the foreseeable future18. Based on a panel of
22 advanced countries between 1955 and 2010, these
authors find a statistically robust correlation between an
economy’s age structure and inflation: the larger the
proportion of dependents – both young and old – the
higher inflation goes, whereas lower inflation is
associated with a larger share of people of working
age19. Juselius and Takats reckon that demography
explains about a third of the variation in inflation and the
bulk of the slowdown in inflation that occurred between
the late 1970s and early 1990s.
Their explanation, albeit tentative, is straightforward.
People’s economic needs and contributions vary over
the various stages of life. Specifically, the ratio of
consumption to production tends to be high for youth
and the elderly and low for working-age adults. As
children and the elderly consume more goods and
services than they produce, a higher share of
dependent people, young and old, should be expected
to lead to excess demand and thus to inflationary
tendencies. By contrast, working people tend to
consume less than they produce, both through their
work and through their savings (which support capital
accumulation), so a higher share of the working-age
population will supposedly increase the supply of goods
and services relative to demand, leading to excess
supply and thus to a deflationary bias. Accordingly,
Juselius and Takats expect that rising old-age
dependency ratios will lead to significant inflationary
pressures in coming decades.
FOR SOME, AGING WILL CREATE A CHALLENGING
INFLATIONARY
ENVIRONMENT
IN
THE
COMING
YEARS
18
Rising ratio of dependents to workers
Juselius and Takats, in a recent paper for the Bank for
International Settlements, argue that the sharp rise in
the dependency ratio will likely foster faster rates of
See Mikael Juselius and Elod Takats (2016), “The agestructure-inflation puzzle”, Bank of Finland Research Discussion
Paper, March 1, and Mikael Juselius and Elod Takats (2015),
“Can demography affect inflation and monetary policy?”, BIS
Working Papers, N°485, February. Also see Mikael Juselius and
Elod Takats (2016), “Age and inflation”, Finance & Development,
Vol.53, N°1, March.
19
17
See, for example, Bloom, David E., David Canning, and Gunther
Fink (2011), “Implications of Population Aging for Economic
Growth”, PGDA Working Paper N°64, January.
For instance, these authors find that in the USA the baby
boomers boosted inflation by 6 percentage points between 1955
and 1975 and lowered it by 5 percentage points between 1975
and 1990, when they entered working age.
9
ECONOTE | N°35 – NOVEMBER 2016
Declining workforce growth
All things being equal, population aging ought to lead to
declining labor force supply20, which, in turn, should be
expected to pave the way for an improvement in the
bargaining power of labor that could drive wages up
relative to the remuneration of other factors of
production, such as capital and land. This is the
mechanism emphasized by Goodhart, Pradhan and
Pardeshi (2015)21. Goodhart et al. argue forcefully that
the demographic "sweet spot" (or the high share of
working-age people relative to the total population) that
has shaped the framework of the global economy over
the past four decades, and which was made even
sweeter by the entry of China and the former Soviet
bloc into the global trading system, is now reversing,
spelling the end of a multi-decade period of cheap labor
and downward pressure on inflation.
Over the past four decades, globalization has led to an
unusually ample global labor supply, which has caused
labor power to collapse in the west, pushing wages
down and creating substantial disinflationary pressures.
But this oversupply of labor is now going into reverse,
Goodhart et al. argue, given that the share of the
working-age population is starting to fall while the entry
of China into the world economy was a one-off supply
20
Labor force growth is driven by a combination of changes in the
overall labor force participation rate (the proportion of the
population in the labor force) and changes in population. The labor
force participation rate is typically lower for 16-to-19-year-olds in
comparison to other age groups; it increases during the prime
working years and then declines sharply after age 55 as workers
leave the labor force. In most advanced countries, the labor
market is now experiencing a negative demographic effect in
which a large segment of the population is moving from an age
group with higher participation rates to an older age group with
lower participation rates, hence, a decline in labor force
participation rates which causes the growth of the labor force to
slow down.
21
See Goodhart, Charles, Manoj Pradhan and Pratyancha
Pardeshi (2015), “Could Demographics Reverse Three MultiDecade Trends?”, Global Issues, Morgan Stanley, September 15.
Also see Charles Goodhart, Philipp Erfurth (2014), “Demography
and economics: Look past the past”, Vox, 4 November, and Alan
Greenspan (2007), The Age of Turbulence: Adventures in a New
World, Penguin Press.
shock. As the positive labor supply shock of the last
decades turns into a negative shock, they argue, the
bargaining power of labor is bound to improve, pushing
up real wages and causing the disinflationary trend of
the three past decades to end and ultimately reverse.
Goodhart et al.’s predictions may prove correct, but
they stand in contrast to the experience of Japan which
has faced many years of falling prices despite its rapidly
aging population and declining workforce. For Goodhart
et al., the lack of inflation in Japan is related to very
specific circumstances. When Japan’s demographics
turned in the early 1970s, they explain, its neighbors in
Asia were still benefitting from a population “sweet
spot”, with a marked expansion in their workforce. This
ample labor supply abroad, they believe, led Japanese
companies to offshore their productions, preventing
local wages from rising. But today, they argue, the
picture is very different, as most of the world is now at
the point where workforce growth is slowing, which
suggests that labor will become increasingly scarce
worldwide, which should inevitably lead to a bidding war
for workers and, thereby, to wage-generated inflation
across the developed world.
What happens next to the bargaining power of labor is
clearly very important for price dynamics. Yet it is worth
remembering a couple of points. First, there probably
remains a reserve army of global workers, as there is
still a large dormant segment of the labor force in China
(where millions of rural and urban residents are
unemployed or underemployed) and a vast supply of
cheap, non-Chinese labor. Second, it might be the case
that the integration of China and other emerging
countries such as India into the world economy has led
to a fundamental change in the bargaining power of
labor in the west. In particular, the waning power of
trade unions across much of the developed world and
the trend towards rising labor market flexibility are
unlikely to go into reverse, which should not be
conducive to sustained wage increases in coming
years. What is more, the future decline in the size of the
labor force may not be as sharp as currently
anticipated, as current projections of the future labor
force neglect behavioral change and institutional
10
ECONOTE | N°35 – NOVEMBER 2016
response
to
changing
age
structures
and
life
expectancies (see Box 3).
BOX 3. DEMOGRAPHY IS (NOT QUITE) DESTINY
As labor supply is higher among working-age adults than among the elderly or youth, a country with large and
rising cohorts of elderly should expect a slowdown in the growth rate of its future workforce or an outright
contraction of its workforce. Actual labor force growth is, in effect, falling off throughout the developed world,
with labor force shrinkage already under way in Japan and in parts of Europe.
There are nevertheless a number of potential antidotes to labor-market tightening stemming from population
22
aging . Specifically, policymakers can mitigate the impact of aging by:




23
Lifting the participation rate of more economically-inactive persons, notably women and older people
Lengthening working life
Raising net immigration
Increasing investment in education and health.
Female labor-force participation
Greater female labor-force participation could contribute to expanding the effective size of the labor force.
Bloom et al. (2009), for instance, find that average female labor force participation tends to rise when the
24
number of children per woman declines . As current population projections predict substantial declines in
total fertility rates over the next decades, sizeable increases in female labor force participation can be
expected.
At present, in the OECD area as a whole, 63% of women aged 15 to 64 are in the labor force (67% in France,
Japan and the USA and 73% in Germany), and the OECD average for 65 to 69-year-olds is just 20% (5% in
France, 11% in Germany, 28% in the USA and 32% in Japan), compared with 32% for men (7% in France,
19% in Germany, 37% in the USA and 54% in Japan). Women remain a substantial reservoir of potential
22
See, for example, Börsch-Supan, Axel, Klaus Härtl and Alexander Ludwig. 2014. "Aging in Europe: Reforms, International Diversification,
and Behavioral Reactions." American Economic Review, 104(5): 224-29.
23
The labor force participation rate is defined as the percentage of economically active persons (i.e. employed plus unemployed) in the
working age population. It thus describes their general propensity to make themselves available to the labor market.
24
See Bloom, D. E., D. Canning, G. Fink and J. E. Finlay (2009), “Fertility, Female Labor Force Participation, and the Demographic
Dividend”, Journal of Economic Growth 14(2): 79-101.
11
ECONOTE | N°35 – NOVEMBER 2016
workers. Importantly, though, suitable work arrangements for women must be implemented to avoid that an
increase in women’s participation rate further depresses the birth rate.
Elder workforce participation
Other priority areas of focus for policymakers should be to encourage the participation rate of older people
and to stretch working life. In the OECD area as a whole, only 50% of 60 to 64-year-olds (both men and
women) are still employed (30% in France, 56% in Germany, 55% in the USA and 65% in Japan), while the
figure for 65 to 69-year-olds is a mere 26% (6% in France, 15% in Germany, 32% in the USA and 42% in
Japan).
Given rising life expectancy and improved health outcomes for the elderly, people may choose to remain in
25
the workforce for longer . More flexible old-age pension arrangements combined with adjustments in tax and
26
benefit policies could change incentives and encourage a higher participation rate among “young” seniors .
And increasing the legal age of retirement would effectively force people to work longer before retiring.
With a 42% share of workers aged 65-69, Japan shows that there is plenty of potential for workers in Europe
and in the USA to retire later. Projected declines in the ratio of the working-age to non-working-age
populations are much less sharp if, instead of defining “working age” to be 15 to 64 years of age, one takes
70 years as the upper age limit for employment (a development that improved standards of living and
medicine should facilitate) to reflect the prospect of later retirement.
Inward immigration
More inward migration could also mitigate labor-market tightening due to demographics in countries with “old”
populations. In 2015, Europe, and most notably Germany, was the scene of one of the greatest crosscontinent mass migrations in recent history, with an influx of more than 1 million Syrians. However, looking
forward, social and political opposition to sustained mass immigration (because of fears of social tensions
and political backlash) will likely prevent increased immigration on a scale large enough to counteract the
27
potential economic effects of population aging in most advanced countries .
Investment in education and health
Finally, increasing investment in education and health could help to counter the declining labor supply that will
come about because of demographics. Prettner et al. (2012), for example, find that declining fertility goes
hand in hand with increasing human capital endowment per person in terms of education and health, which,
28
in turn, could lead to a larger and more productive effective labor force in the future . This strand of research
builds upon the idea that economic growth is not merely a function of the quantity of labor but also of its
quality, which is increased by greater education and health investments.
All in all, changes in individual behavior and in the policy framework are set to mitigate the negative impact of
aging on the future size of the workforce, but these anticipated changes are nevertheless unlikely to be able
to completely reverse the overall declining labor force trend that current projections of population size and
structure predict for the decades to come in the developed world.
25
See, for example, Bloom, D. E., D. Canning, G. Fink and J.E. Finlay (2007), “Demographic Change, Institutional Settings, and Labor
Supply”, mimeo. These authors show that the theoretically optimal response to rising life expectancy is to increase working life
proportionately, without increasing savings rates. However, this theoretical response has so far largely failed to materialize owing to various
factors including difficult labor markets and social security systems. In many countries, the latter provide strong incentives to retire rather
than to continue working. In practice today, penalties for working past the national retirement age, coupled with increases in uncertainty
about the availability of benefits often lead individuals to save more to finance consumption in retirement.
26
See, for example, Imrohoroğlu, Selahattin and Sagiri Kitao. 2012. "Social Security Reforms: Benefit Claiming, Labor Force Participation,
and Long-Run Sustainability." American Economic Journal: Macroeconomics, 4(3):96-127; and Sánchez-Romero, Miguel, Jože Sambt, and
Alexia Prskawetz (2013), “Quantifying the Role of Alternative Pension Reforms on the Austrian Economy”, Labor Economics, Volume 22,
June, pp. 94-114. Importantly, Sánchez-Romero and al. find that an increase in the educational distribution of the work force would
increase the incentive for later retirement ages and thereby help counter the declining labor supply stemming from lower fertility.
27
Stabilizing the share of elderly people in the populations of advanced countries would require an immediate eightfold increase in
immigration from less-developed countries, according to the International Monetary Fund.
28
See Prettner, Klaus, David E. Bloom, and Holger Strulik (2012), “Declining Fertility and Economic Well-Being: Do Education and Health
Ride to the Rescue?”, IZA Discussion Paper No. 6527, April. Based on a panel of 118 countries between 1980 and 2005, these authors
show that more human capital investment can partly offset the negative impact of declining fertility on the effective size of the labor force.
Also see e.g. Ahituv, A. (2001), “Be Fruitful or Multiply: On the Interplay between Fertility and Economic Development”, Journal of
Population Economics, Vol. 14:51-71; Bloom, D. E. and Canning, D. (2000), “The Health and Wealth of Nations”, Science, Vol. 287:12071209; Lee, R. and A. Mason (2010), “Fertility, Human Capital and Economic Growth over the Demographic Transition”, European Journal
of Population 26: 159-182.
12
ECONOTE | N°35 – NOVEMBER 2016
FOR
OTHERS,
A
DECLINING
AND
AGING
POPULATION IS INHERENTLY DEFLATIONARY
Loss of human capital
According to other authors, the impact of ongoing
demographic shifts will be mostly deflationary. Fujita
and Fujiwara (2014)29 developed a model with
demographic and skill transitions, and specialized
knowledge (that is, skills and abilities in one particular
area). In their model, the skill the worker possesses is
firm specific, which means that if the worker loses his
job, he also loses his skills. These authors examine the
ebb and flow of productivity in the course of the
demographic transition. Initially, the decline in fertility
causes the share of older and, thus, more-experienced
and more-productive workers in the labor force to
increase, leading to increased output and real wages,
and higher inflation.
Eventually, however, as labor force aging progresses,
firm specificity (or specialized knowledge) renders the
older workers increasingly vulnerable to human capital
depreciation. This is because when older workers quit
their jobs, they lose their human capital, and then
become less productive30. So it is held that in the late
stage of labor force aging, a higher share of older
workers implies a higher share of inexperienced
workers (because of the skill depreciation at the time of
job loss), hence, a fall in nation-wide labor productivity
and in real wages. And lower real wages, in turn, put
downward pressure on inflation. When the model is
subject to a significant rise in the pace of job losses,
such as the one experienced in Japan in the 1990s and
29
See Fujita, Shigeru, and Ippei Fujiwara (2014), “Aging and
Deflation: Japanese Experience”, Unpublished manuscript, July.
http://conference.nber.org/confer/2014/JPMs14/Fujita_Fuji
wara.pdf.
30
Several researchers argue that workforce aging has negative
implications for labor productivity owing to factors such as the
depreciation of knowledge, lessened physical and cognitive
abilities and lower effectiveness of adaptation to new
technologies. See, for example, International Monetary Fund
(2016), “The Impact of Workforce Aging on Euro Area
productivity”, IMF Country Report No. 16/220, July; Dixon, Sylvia
(2003), “Implications of Population Aging for the Labor Market”,
Labor Market Trends, February. On the other hand, the
accumulation of experience over time tends to increase
productivity. See, for example, Disney, Richard (1996), “Can we
Afford to Grow Older? A Perspective on the Economics of Aging”,
MIT Press, Cambridge: Mass. Arguably, the impact of workforce
aging on productivity will differ from sector to sector, with a likely
negative effect in basic jobs (such as tillers) but a positive effect in
such professional areas as medicine or law, in which the
accumulation of knowledge and experience may have a favorable
influence on efficiency. It is a key issue for researchers to analyze
whether a smaller, but older, labor force in the future can be as
productive as the previous generations. Another factor to consider
when assessing the prospects for productivity in an aging world is
that each new cohort of workers tends on average to be better
educated than the retiring cohort as a more educated labor force
is more productive than a less educated labor force.
2000s, the mechanism in the model leads to higher
levels of deflation. Likewise, a greater contraction in the
labor force size increases the magnitude of deflation.
Lower expectations of future economic growth
Former Bank of Japan Governor Masaaki Shirakawa,
has long claimed that a declining and aging population
will lead to deflationary pressure. To quote him: “A
cross-country comparison among advanced economies
reveals intriguing evidence: Over the decade of the
2000s, the population growth rate and inflation correlate
positively across 24 advanced economies […] A closer
look at the case of Japan confirms the increasingly
positive correlation between inflation and population
growth since the 1990s” [see Shirakawa (2012)]31.
According to Shirakawa, the key to the link between
population aging and deflation is reduced expectations
of future earnings. It is well recognized that a decline in
the growth of the labor force will reduce a country’s
capacity to grow32, which should raise concerns about a
future slowing of economic growth and thus about
future earnings.
Shirakawa believes that lower expectations of future
earnings associated with a declining and aging
population will lead to deflationary pressure by
prompting forward-looking households to consume less
and save more today. Moreover, he emphasizes that
the public generally tends to underestimate the negative
impact of demographic changes which has the potential
to magnify the downward pressure on prices. As public
awareness of yet-to-materialize declines in GDP growth
slowly builds up, he argues, forward-looking responses
will likely come into play, precipitating current falls in
aggregate demand, with their attendant deflationary
effects.
Demography as a source of underestimated
shocks
Katagiri (2012) also stresses the implications of a
faster-than-expected aging phenomenon for price
dynamics33. This author argues that the repeated
upward revisions in the official forecast for the speed of
Japanese population aging have worked as unexpected
31
See, for example, Masaaki Shirakawa (2012), “Demographic
Changes and Macroeconomic Performance – Japanese
Experiences”, Opening remarks by Mr Masaaki Shirakawa at the
2012 BOJ-IMES Conference, hosted by the Institute for Monetary
and Economic Studies, at the Bank of Japan, Tokyo, 30 May.
32
The impact on aggregate real GDP of a declining and aging
population is straightforward given the direct effect on the size of
labor inputs.
33
See Mitsuru Katagiri (2012), “Economic Consequences of
Population Aging in Japan: Effects through Changes in Demand
Structure”, Discussion Paper N°2012-E-3, Institute for Monetary
and Economic Studies, Bank of Japan, March. Katagiri notes that
the forecast is revised upward every time officials make a new
forecast.
13
ECONOTE | N°35 – NOVEMBER 2016
demand shocks leading to deflationary pressure of
about 0.3 percentage points a year. Katagiri’s reasoning
is as follows: young people, he observes, tend to buy
more manufacturing goods (like cars and electric
appliances) than old people who spend more for
services such as medical care and tourism. It follows
that population aging should lead to a demand shift
from manufacturing sector products to the services
sector which may, in turn, bring about a (re)allocation of
resources (labor output) from the manufacturing sector
to the services sector34. However, the productivity of the
services sector is typically lower than that of the
manufacturing sector.
As a result, the argument goes, aging countries will
likely experience significant resource reallocation from
high productivity sectors to low productivity ones,
leading to a decline in aggregate productivity. Lower
aggregate productivity, however, lowers the trend
growth rate of potential GDP which, according to
standard economic theory, should push down the level
of the natural rate of interest35. Katagiri argues that,
because population aging in Japan has continuously
surprised on the upside, the central bank may not have
relaxed
its
monetary
stance
sufficiently
to
accommodate the decline in the natural rate of interest,
which has opened up a negative output gap that has
fueled deflationary pressure.
Changing consumption preferences
According to the life-cycle theory developed by Franco
Modigliani, individuals smooth their consumption over
their lifetimes, buying housing and going into debt when
young, saving and buying assets during their adult lives
and running down their savings and selling their assets
once they retire36. So it is often held that household
34
On the impact of aging on the supply-side of the economy, with
the expansion of the service sector relative to manufacturing, see,
for example, Siliverstovs, Boriss, Kholodilin, Konstantin, Thiessen,
Ulrich (2011), “Does Aging Influence Structural Change? Evidence
from panel data, Economic Systems, Elsevier, vol. 35(2), pp. 244260, June.
35
The natural rate of interest – a concept which was first proposed
by Wicksell in 1898 – is the real, or inflation-adjusted, interest rate
that is consistent with an economy at full employment and with
stable inflation. If the effective real rate of interest is above (below)
the natural rate then monetary conditions are tight (loose), leading
to underutilization (overutilization) of resources and thus lower
inflation. Lower potential growth can affect the natural rate of
interest via two channels. First, it lowers the return on capital and
thus leads to lower investment demand. Second, it lowers
households’ permanent income and thus encourages forwardlooking households to consume less and save more. Lower
investment demand combined with higher savings pushes down
the natural rate of interest.
36
See Modigliani, Franco and Richard Brumberg (1954), “Utility
Analysis and the Consumption Function: An interpretation of
Cross-section Data”, In Post-Keynesian Economics, ed. Kenneth
K. Kurihara, 388-436. New Brunswick, N.J.: Rutgers University
Press. Also see Modigliani, F. and Ando, A. (1957), “Test of the
Life-cycle Hypothesis of Savings”, Bulletin of the Oxford University
demand for housing services and housing capital (as a
conduit of saving) goes through a life cycle, peaking
during people’s working life, as individuals start and
maintain a family, and then declining when they retire,
as they move to smaller houses or peripheral locations
or shift from homeownership to renting37. If true, this
should be reflected in changes in housing and land
prices. Using the IMF’s Global Integrated Monetary and
Fiscal (GIMF) model, Anderson, Botman and Hunt
(2014) find that, in Japan, an aging and declining
population created substantial deflationary pressures
mainly though a fall in land prices and declining
growth38. They also show that the repatriation of huge
amounts of foreign assets by Japanese retirees to pay
for their consumption in retirement led to real exchange
rate appreciation, which added to deflationary pressure
by making imports cheaper. Finally, they stress that, in
Japan, the deflationary effects of aging were magnified
by the large fiscal consolidation need.
Based on a panel dataset covering 30 OECD
economies for the period 1960-2013, Yoon, Kim and
Lee (2014) also find that an aging or declining
population is significantly deflationary39. These authors
argue that, while an aging or declining population tends
to increase inflationary pressure by reducing the
effective supply of labor in the economy, it also leads to
multifarious deflationary demand-side effects through
notably lower consumption, changing relative prices
reflecting different consumption preferences40, and a
negative wealth effect from falling asset prices. While
emphasizing that the specific channels through which
aging affects the inflation rate are yet to be examined,
Yoon and al. argue that the deflationary effects of an
aging population should dominate the inflationary
Institute of Economics and Statistics 19, pp.99-124. See Franco
Modigliani (1970), “The Life-Cycle Hypothesis and Inter-country
Differences in the Saving Ratio”, pp. 197-225 in W. A. Eltis, M.
FG. Scott, and J. N. Wolfe, eds., “Induction, Growth, and Trade:
Essays in Honor of Sir Roy Harrod”, (Oxford University Press).
37
See, for example, Pitkin, J.R and D. Myers (1994), “The
Specification of Demographic Effects on Housing Demand:
Avoiding the Age-Cohort Fallacy”, Journal of Housing Economics,
3(3): 240-250; Flavin, T. and Yamashita (2002), “Owner-Occupied
Housing and the Composition of the Household Portfolio”,
American Economic Review, 92(1): 345-362; Kraft H. and C. Munk
(2011), “Optimal Housing, Consumption, and Investment
Decisions over the Life Cycle”, Management Science, 57(6): 10251041.
38
Anderson, Derek, Dennis Botman, and Ben Hunt (2014), “Is
Japan’s Population Aging Deflationary?”, IMF Working Paper,
WP/14/139, August.
39
See Yoon, J. W., Kim, J., and J. Lee (2014), “Impact of
Demographic Changes on Inflation and the Macroeconomy”, IMF
Working Paper WP/14/210.
40
Typically, young people tend to spend more on housing,
transportation, communication, and education, while old people
tend to spend more on services expenditures such as medical
care and travel.
14
ECONOTE | N°35 – NOVEMBER 2016
effects because aggregate supply will likely adjust at a
slower pace than aggregate demand.
Lower desire to invest and to save
In the late 1930s, Alvin Hansen argued that the decline
in the birth rate in the United States was a major cause
of the shortfall in investment and hence the lack of
aggregate demand which characterized the Great
Depression41. Firms need a given capital stock
(equipment, structures and land) per worker. So if
population growth slows, the demand for new houses,
new office buildings and new capital goods to equip
new workers falls. Hansen argued at the time that
“secular stagnation”42 was threatening the US. Slowing
population growth, he emphasized, implies an enduring
decline in investment demand, which can create a
chronic oversupply of savings that can push the
economy into a semi-permanent slump with associated
deflationary pressures. The (unexpected) post-World
War II baby boom altered the population dynamics in
the USA and rebutted Hansen’s predictions, but the
current demographic landscape gives renewed
relevance to the secular stagnation hypothesis, as
argued by Summers (2016)43 and others. Secular
stagnation will arise if desired investment remains
persistently below desired saving, creating chronically
deficient aggregate demand and hence deflationary
pressure.
So a critical question for future price dynamics is
whether slowing worker and population growth, or an
outright decline, will reduce the rate of investment by
more than it reduces saving. There is little dispute
among economists that a lower rate of population
growth will reduce the demand for new investment while
lowering the potential growth rate of the economy. At
41
See Hansen, Alvin (1939) "Economic progress and declining
population growth", American Economic Review, 29(1): 1-15.
42
The term “secular stagnation” was coined by Hansen in his 1938
address to the American Economic Association.
43
See Lawrence H. Summers (2013), “Speech given at the IMF
Fourteenth Annual Research Conference in Honor of Stanley
Fischer”, November 8. Also see Lawrence H. Summers (2016),
“The Age of Secular Stagnation: What It Is and What to Do About
It,” Foreign Affairs, February.
the same time, however, many economists anticipate
that as populations grow older, aggregate saving will
eventually decline as a percentage of national income44.
This is what the widely accepted life-cycle theory of
saving predicts45. The life-cycle theory holds that the
savings pattern of households tends to follow a path
that changes as households age, with savings first
rising, as working households increase their provision
for retirement, then reaching a lifetime peak when
workers are at the middle and near the end of their
careers, and ultimately falling as these households
move through retirement and begin to dissave to live off
accumulated assets. So, according to the life-cycle
theory, countries undergoing aging should eventually
expect a fall in aggregate savings as an increasingly
large share of households move through into retirement
and begin to dissave.
It is on this basis that Goodhart and Erfuth (2014)
predict that population aging will cause savings rates to
fall and real interest rates to head back up46. Dissaving
by retirees, in turn, could start to put upward pressure
on inflation as this should support aggregate demand at
a time when aggregate supply is declining owing to a
declining workforce. Goodhart and Erfuth expect that,
as aging progresses, excess demand, owing to
declining aggregate saving, will combine with rising
wages, due to competitive bidding for increasingly scare
labor, to create sustained inflationary pressure. For
these authors, the ex-ante desire to invest should fall
somewhat as the population ages, but it should fall less
than the decline in the ex-ante desire to save – hence,
44
See, for example, Bosworth, P. Barry, Ralph C. Bryant, and
Gary Burtless (2004), “The Impact of Aging on Financial Markets
and the Economy: A Survey”, Brooking Institution Working Paper
(Washington: Brookings Institution); Mc Morrow, K., and Roeger,
W. (2004), “The Economic and Financial Market Consequences of
Global Aging”, Journal of Economics, Volume 83, Issue 2, pp.
202-204.
45
See Modigliani, Franco and Richard Brumberg (1954) (op. cit.).
46
Op.cit. Goodhart and Erfurth (2014) argue that by 2025 real
interest rates should have returned to the historical equilibrium
value of around 2.5–3%, with nominal rates at 4.5–5%.
15
ECONOTE | N°35 – NOVEMBER 2016
their conclusion that real rates of interest will go back
up47.
Yet whether the aggregate saving ratio will rise or
decline when the population ages is a moot point.
Based on a panel of 6 OECD nations48 between 1970
and 1991, Poterba (1994), for instance, finds little
evidence supporting the life-cycle hypothesis49. In
particular, he finds that virtually everywhere the elderly
do not dissave to finance their consumption during
retirement (possibly because of a desire for intergenerational transmission of wealth), and that young
families in their twenties and thirties save a positive and
increasing proportion of their income. Both observations
stand in sharp contradiction to the hump-shaped pattern
of saving predicted by the life-cycle theory. In general,
the empirical microeconomic evidence that has
emerged in the past twenty years tends to suggest that
the elderly do not dissave as much as the original
version of the life-cycle theory predicts they should50.
Some argue that aggregate saving could even rise as
countries undergo aging. Bloom et al. (2003, 2009), for
example, find that increased life expectancy – and thus
longer anticipated periods of retirement – is generally
associated with higher savings rates51. Higher adult
survival and the anticipation of longer periods of
retirement, coupled with precarious prospects for social
security payments will likely induce individuals to
increase savings over their working life (especially in
countries where policies and institutions deter people
from working past their early or mid-60s) in order to
finance a continued high standard of living during
retirement. This may also induce them to remain in the
workforce for longer and draw down savings at a later
stage. Skill and education is also set to play a role as
better-educated older people are far more likely to work
for longer and to accumulate more savings and assets
than they will draw down during retirement. All in all,
increased longevity, pension reform and a rising share
of well-educated elderly are likely to be drivers of higher
desired saving as people will want to prepare for a
longer prospective retirement period.
At this stage in the research, the question as to whether
slower workforce growth will reduce the desire to invest
by more than it reduces the desire to save remains
largely undetermined. If older people were to work
longer and prefer to carry higher levels of saving into
retirement than they have done to date at a time when
investment intentions are falling due to slowing growth
in the labor force, a condition of the sort envisaged by
Hansen, with persistently low growth, underemployment
of resources and deflationary pressure, could clearly
come about.
Stronger constituency to keep inflation low
47
Goodhart and Erfuth (2014) predict that as the cost of labor
rises relative to capital owing to declining availability of labor, the
capital-to-labor ratio will rise again.
48
Namely, Canada, Germany, Italy, Japan, the UK and the USA.
49
Poterba, James M. (1994), “International Comparisons of
Household Saving”, Chicago, London: University of Chicago
Press.
50
There is a very substantial difference between the results of
studies based on macroeconomic or cross-national panel
evidence (which usually confirm the predictions of the life cycle
model) and those based on microeconomic evidence (which often
find no, or only modest, effects on savings of aging populations).
These differences stem at least in part from the use of different
types of data and methods. On this point see, for example,
Bosworth et al. (2004) (op.cit.). These authors note that the
distribution of wealth and saving across households tends to be
highly skewed - notably in the USA which has been the object of
many studies -, and that the small percentage of households that
accounts for an out-size fraction of private saving is poorly
represented in most microeconomic studies. However, they
emphasize, the saving behavior of the most-affluent households
will likely differ significantly from that of average households (i.e.
the saving behavior of median households will most probably differ
from that of average households), which could explain, at least in
part, the differences in results between the macroeconomic
studies and the microeconomic analyses.
Bullard, Garriga, and Waller (2012) resort to a politicoeconomic hypothesis to argue that population aging can
put downward pressure on inflation or even lead to
deflation52. Typically, the elderly are richer in financial
and real capital while the young are richer in human
capital. So, while the former will tend to depend more
on the return of their assets, the former will rely on
wages as their main source of income. As older
societies will typically have higher average asset and
wealth holdings, a graying society should be expected
to have higher concern for, and hence, aversion to
inflation. In contrast, societies dominated by young
51
See Bloom, D. E., D. Canning and B. Graham (2003),
“Longevity and Life-Cycle Savings”, Scandinavian Journal of
Economics 105(3): 319-38. Also see Bloom, D. E., Canning, D.,
Mansfield, R. K., and Moore, M. (2007), “Demographic Change,
Social Security Systems, and Savings”, Journal of Monetary
Economics, Volume 54, pps.92-114.
52
See Bullard, James, Carlos Garriga, and Christopher J. Waller
(2012), “Demographics, Redistribution, and Optimal Inflation”,
Federal Reserve Bank of St. Louis Review, November/December,
Vol. 94, No. 6, pp. 419-39.
16
ECONOTE | N°35 – NOVEMBER 2016
households will prefer relatively high wages and high
inflation since they do not initially have any assets and
are generally borrowers. Rising aversion to inflation
because of population aging, Bullard et al. argue, could
affect a central bank’s perception of what its objection
function should be and lead to a lower optimal inflation
target. As central banks may be inclined to cater to the
inflation preferences of dominant age groups, they
could employ monetary policy more aggressively to
combat inflation. If true, more elderly societies should
be associated with lower inflation.
In a similar vein, Katagiri, Konishi and Ueda (2014)
suggest that government policies may reflect voter
preferences. These authors establish a distinction
between aging caused by an increase in longevity and
that produced by a decline in the birth rate53. Increased
longevity, they argue, causes the ranks of pensioners to
rise and thus the political influence of the older
generation to increase, which may impel the
government to embrace fiscal austerity by raising its
income tax rates (thus, harming the younger
generation) in order to prevent inflation from eating into
the value of the older generation’s assets. In contrast,
when population aging is caused by a decline in fertility,
which is mainly reflected in a falling tax base, a shortlived government may well be tempted to monetize its
debt, pushing inflation up, hence benefiting the young at
the expense of the older generation. These authors find
that in Japan the deflationary effect of higher longevity
overwhelms the inflationary effect of falling fertility. This
is partly because, they believe, the main surprise in
Japanese demography has been ever-increasing
longevity. The authors' simulations of the model show
that population aging in Japan depressed inflation by
about 0.6 percentage points annually over the past 40
years.
A
MIXED
INFLUENCE
VIA
FISCAL
Anderson et al. (2014)]54. But other governments may
resort to money printing, which may prove, by contrast,
inflationary. So depending on the policy response by
governments, we might expect that demographic shifts
will have inflationary or deflationary consequences.
ENTERING UNCHARTERED TERRITORY
The analysis of the effects of expected population aging
on inflation represents unchartered waters, partly
because
the
ongoing
demographic
shift
is
unprecedented in world history, both in its nature and its
magnitude. Only Japan has so far entered into this
challenging new world. Assessing the applicability of
Japan’s deflationary experience to other advanced
economies is clearly a key issue for researchers. The
influence of demographic aging on inflation is bound to
become a high-profile topic in the public debate in the
years ahead. As for the theoretical debate, it is far from
over.
POLICY
VARIABLES
Graying populations will inevitably exert upward
pressure on public budgets. Population aging will
deteriorate fiscal balances while leading to greater
spending in such areas as pensions, health care and
other welfare programs for the elderly, at a time when
the falling share of the productive labor force will slow
economic growth, thereby reducing the tax base. To
cope with these mounting fiscal pressures, some
governments may choose to implement fiscal
consolidation measures (possibly because of the
increased political influence of the elderly) which would
cause deflationary pressures by leading to slow growth
and an increased output gap [see, for example,
54
53
See Katagiri, Hideki Konishi, and Kozo Ueda (2014), “Aging and
Deflation from a Fiscal Perspective”, WINPEC Working Paper
Series No. E1413, November.
Anderson, Derek, Dennis Botman, and Ben Hunt (2014), op.cit.
These authors suggest that the deterioration in fiscal balances
brought on by population aging, coming on top of elevated initial
deficit and debt burdens, can lead to a rising risk premium and/or
fiscal consolidation which can cause aggregate output to fall below
potential, thus fueling deflationary pressure.
17
ECONOTE | N°35 – NOVEMBER 2016
PREVIOUS ISSUES OF ECONOTE
N°34 Emerging markets’ external debt: It’s the same old song?
Juan Carlos DIAZ MENDOZA (November 2016)
N°33 US public debt: Towards more domestic and private financing
Amine TAZI, Clémentine GALLÈS (September 2016)
N°32 China: Assessing the global impact of a Chinese slowdown
Sopanha SA, Théodore RENAULT (July 2016)
N°31 France: A private sector in better financial health, despite higher debt
François LETONDU (June 2016)
N°30 A world without inflation
Marie-Hélène DUPRAT (March 2016)
N°29 Low interest rates: the ‘new normal’?
Marie-Hélène DUPRAT (September 2015)
N°28 Euro zone: in the ‘grip of secular stagnation’?
Marie-Hélène DUPRAT (March 2015)
N°27 Emerging oil producing countries: Which are the most vulnerable to the decline in oil prices?
Régis GALLAND (February 2015)
N°26 Germany: Not a “bazaar” but a factory!
Benoît HEITZ (January 2015)
N°25 Eurozone: is the crisis over?
Marie-Hélène DUPRAT (September 2014)
N°24 Eurozone: corporate financing via market: an uneven development within the eurozone
Clémentine GALLÈS, Antoine VALLAS (May 2014)
N°23 Ireland: The aid plan is ending - Now what?
Benoît HEITZ (January 2014)
N°22 The euro zone: Falling into a liquidity trap?
Marie-Hélène DUPRAT (November 2013)
N°21 Rising public debt in Japan: how far is too far?
Audrey GASTEUIL (November 2013)
N°20 Netherlands: at the periphery of core countries
Benoît HEITZ (September 2013)
N°19 US: Becoming a LNG exporter
Marc-Antoine COLLARD (June 2013)
N°18 France: Why has the current account balance deteriorated for more than 20 years?
Benoît HEITZ (June 2013)
N°17 US energy independence
Marc-Antoine COLLARD (May 2013)
N°16 Developed countries: who holds public debt?
Audrey GASTEUIL-ROUGIER (April 2013)
N°15 China: The growth debate
Olivier DE BOYSSON, Sopanha SA (April 2013)
18
ECONOTE | N°35 – NOVEMBER 2016
19
ECONOTE | N°35 – NOVEMBER 2016
ECONOMIC STUDIES
CONTACTS
Olivier GARNIER
Group Chief Economist
+33 1 42 14 88 16
[email protected]
Aurélien DUTHOIT
Macro-sectoral analysis
+33 1 58 98 82 18
auré[email protected]
Olivier de BOYSSON
Emerging Markets Chief Economist
+33 1 42 14 41 46
[email protected]
Juan Carlos DIAZ MENDOZA
Latin America
+33 1 57 29 61 77
[email protected]
Marie-Hélène DUPRAT
Senior Advisor to the Chief Economist
+33 1 42 14 16 04
[email protected]
Nikolina NOPHAL BANKOVA
Macro-sectoral analysis
+33 1 58 98 89 09
[email protected]
Ariel EMIRIAN
Macroeconomic analysis / CEI Countries
+33 1 42 13 08 49
[email protected]
Emmanuel PERRAY
Macro-sectoral analysis
+33 1 42 14 09 95
[email protected]
Isabelle AIT EL HOCINE
Assistant
+33 1 42 14 55 56
[email protected]
Clémentine GALLÈS
Macro-sectoral analysis / United States
+33 1 57 29 57 75
[email protected]
Sopanha SA
Asia
+33 1 58 98 76 31
[email protected]
Yolande NARJOU
Assistant
+33 1 42 14 83 29
[email protected]
François LETONDU
Macroeconomic analysis / Euro zone
+33 1 57 29 18 43
[email protected]
Danielle SCHWEISGUTH
Western Europe
+33 1 57 29 63 99
[email protected]
Sigrid MILLEREUX-BEZIAUD
Information specialist
+33 1 42 14 46 45
[email protected]
Constance BOUBLIL-GROH
Central and Eastern Europe
+33 1 58 98 98 69
[email protected]
Mathieu NOGUÈS
Research assistant
+33 1 58 98 79 50
[email protected]
Mathieu NOGUÈS
Publishing
+33 1 58 98 79 50
[email protected]
Société Générale | Economic studies | 75886 PARIS CEDEX 18
http://www.societegenerale.com/en/Our-businesses/economic-studies
Tel: +33 1 42 14 55 56 — Tel: +33 1 42 13 18 88 – Fax: +33 1 42 14 83 29
All opinions and estimations included in the report represent the judgment of the sole Economics Department of Societe Generale and do
not necessary reflect the opinion of the Societe Generale itself or any of its subsidiaries and affiliates. These opinions are subject to change
without notice. It does not constitute a commercial solicitation, a personnal recommendation or take into account the particular investment
objectives, financial situations.
Although the information in this report has been obtained from sources which are known to be reliable, we do not guarantee its accuracy or
completeness. Neither Societe Generale nor its subsidiaries/affiliates accept any responsibility for liability arising from the use of all or any
part of this document.
Societe Generale may both act as a market maker or a broker, and may trade securities issued by issuers mentioned in this report, as well
as derivatives based thereon, for its own account. Societe Generale, including its officers and employees may serve or have served as an
officer, director or in an advisory capacity for any issuer mentioned in this report.
Additional note to readers outside France: The securities that may be discussed in this report, as well as the material itself, may not be
available in every country or to every category of investors.
20